THE DUNHILL WINDOW PROJECT

How the WLMS came to fill the windows of a top West London store for Christmas...

By Paul Joachim

Early in July 2004, Roger Poulet telephoned me. As is usual for him, he was full of the joys of Spring (well, OK then - early Summer) and of enthusiasm. “Look, Paul”, he started, “I have the most fantastic project all set up. We are going to fill the windows of Alfred Dunhill & Co in Jermyn Street (London's poshest retail street, just yards from Piccadilly and a short stroll from Buckingham Palace) with Meccano.”

“Hold hard!” I responded; “who is going to do the filling and where is the Meccano going to come from?”  “We are, and all the Meccano we could possibly need is already owned by Dunhill's President. Richard Dunhill has been a closet collector for most of his life and he has a colossal collection for us to use. All the chaps are dead keen and it only needs to settle a budget and it will be just fantastic”. “And?”

“Well, I'm off to France in a day or so, for the rest of the Summer, so I thought you would like to negotiate the details...”

To be fair to Roger, he had already decided which WLMS members were best suited to the team and also most likely to be available, and had subsequently persuaded Howard Somerville and Colin Davies to join in. He had also got a sketch of the proposed window display from Dunhill's  Global Visual Merchandising Manager (I jest not about the job title). They wanted twin towers, with a sort of bridge across the top and a couple of carousels between them. They were looking for something to move about the towers – “perhaps a crane or a lift or something...” 

The four of us met the next morning, at Roger's place, for a brief council of war. Howard and I would do the main Meccano designing and building; Howard taking a carousel and one tower and I the other tower.  Colin would be chief engineer, checking the structural sensitivities and health and safety issues and be responsible for the bridge. Roger would build to our instructions,  be project leader and general front man. We could start building on September 6th (everyone then being home from summer holidays) and install in the shop on October 11th. Fortunately the WLMS exhibition at St Albans was to be held on the weekend of September 25th, so we would treat that as a dummy run.

So much for the plan...

My job was to build a Paternoster lift in the left hand tower. Fortunately I had ridden on such a machine and got quite familiar with it while at University – a pair provided the transport for the Engineering faculty. As I remembered, the system spent much of its time ringing alarm bells because someone had tripped the sensors designed to avoid mashing passengers between floors: the reason why nowadays you hardly see them at all. I did some experiments with lengths of chain, then with plastic chain, bicycle chain and finally chain made up of strips linked by bolts and threaded pins. The latter was just about stiff enough to carry the load – eight cars each about a foot high, six inches deep and eight inches wide. And the contents of each had to be spotlit with a 20 Watt halogen lamp... One huge piece of luck was a search on the Internet found “Jan” - an engineer who designed and still maintains the last few Paternoster elevators in the United States. Though “Jan” does not wish to be identified, his guidance on the design of the lifts was invaluable.

The right hand tower would be a conventional lift. Howard wanted to build a mechanical sequencer,  I thought that was risky and suggested an electrical sequencer. There are two broad approaches to this control problem. Either one can make one action trigger the next and then that trigger its follower and so on through the whole process, or you can define the sequence externally and hope that the sequencer will allow enough time for each step to complete and without too many gaps between processes. Howard thought about the first process mechanically and I decided I wanted to do it electrically. After some close argument, we compromised on a very simple electro-mechanical sequencer (of the second kind) which would close contacts fed by a split potential low voltage supply controlling the most reliable 12 Volt motors we could lay our hands on (Faulhaber 24V 21:1 integral reduction gearboxes) for the doors and the hoist. Apart from building it too compactly (neatness overriding ease of maintenance – a common theme in this saga) the idea proved sound.

Colin was to build the beam across the top. This would carry two very slowly-rotating turntables, each suspending a bag or an umbrella or something equally simple, and the space in the middle would be filled with a six foot diameter carousel to carry leatherware, with a central contra-rotating platform for a mannequin. “Easy” – we thought – “just a simple pin and peg drive like Peter's big Blocksetting Crane bases...”

We had all done some thinking – Colin had designed a light weight but rigid triangular truss arrangement for the overhead beam, I proposed a composite girder framework, based on the classic Hammerhead Crane model's tower structure, for the towers and Howard had worked out a cantilevered bearing and very neat contra-rotating drive mechanism for the Carousel.

I then went to meet John Abbate:  Dunhill's  Global Visual Merchandising Manager...  Dunhill had been using Meccano in nickel finish to set off the luxurious leather goods they were offering in their 21 shops worldwide. These were made for them in Wales by their shopfitters; Icon. However, when John had suggested something spectacular for the Flagship London store, Icon suggested he find someone “who knew what he was doing”. It just so happened that John discovered his President's lifelong passion, and Richard Dunhill has been steadily disposing of his enormous collection over the last few years through Hugo March, Director of Toys at Christies, the auctioneers. As you probably know, Roger Poulet is a regular client there too – Dinky Toys being his thing, and so the connection was made. But, did we “know what we were doing”? In retrospect we have concluded we would do almost everything differently – but hindsight is a wonderful thing.

The first task was to establish what Dunhills wanted and why. “Easy”, John explained. “The Dunhill brand is one of the very few that is aimed exclusively at men. We want to associate our name with quality and Britishness almost to a nostalgic level, and with fun... fun and games, big boys' toys and a bit of retro style. We want you to stop the traffic!” I discovered, incidentally, that John comes from a long line of Sicilians, sports an MBA and has a monumental mason for a father.

Second task was the budget. Could they afford us? After discussing the budget for each monthly window change we agreed it would not do, not by miles. “However, I pointed out, “after all that work it would be a shame to pull it out after only a month.” “Of course!” rejoined John. “Red and Green are the colours of Christmas. We will launch the New Connaught leather range in November then re-dress for the fun and games promotion for Christmas.” That more than doubled our budget and suddenly we were in real danger of having a real project and a very tight timescale.

I spent a day at Richard Dunhill's house (well, 19 acre estate actually, near the golf course in Denham) with John Abbate, checking out the collection. Richard has only ever attempted one model – SML4. As it never worked, he decided to confine his interest in Meccano to collecting. Even after five years of Christies' sales, Richard's collection is gargantuan. It fills a double double garage and a workshop – originally built to house four vintage Rolls Royces – but that is another story. Not surprisingly, for a collector's collection, we would need hundreds of extra flat girders, long girders and nuts and bolts.

Naturally, we approached Dave Taylor for the parts - but the numbers were so huge that he was in no position to help us in the time available

In the end, Icon sourced the parts from a number of places and sprayed them all Medium Green. However, Icon forgot to de-grease and prime the parts, so we spent the next two months peeling green paint off everything. Once the “contraption” (The nickname was Dave Taylor’s) appeared at the St Alban's exhibition, he became a tower of strength – getting spare parts to us by Express Delivery on several occasions to keep construction going.

Being a Collector's Collection there was a huge surplus of useless parts like braced girders and hinged flat plates, and shortages of useful parts in turn. However, that was no challenge for the assembled Meccanomen; we simply used them! The plating of the lift doors and fronts and the support of the lift guides used an astonishing number of these parts. Michael Denny's GROPE (Get Rid Of Parts Effectively) process has never been so exploited!

We established ourselves in the garage at Denham. Richard and his charming Argentinian wife, Pat,

looked after us with a generosity far exceeding anything we could have imagined or reasonably expected. At first, while there was not much to show and we were busy building sub-modules, Richard was reticent about helping us find parts. However, once we had been going about a week and the main tower and carousel frames started to take shape, the parts suddenly appeared and Richard regularly ensconced himself in a convenient corner, muttering “remarkable, amazing, quite extra-ordinary!” From this time on, Richard would always greet our arrival with “... and how are my little Meccano boys?”

The household dogs (a huge Great Dane and a hardly any smaller Lurcher) also “helped” by slobbering over everything and appearing at one's elbow just as a critical bolt had to be tightened. However, the introduction of Roger’s little Welsh Terrier, “Mij”, provided an effective security service by keeping them occupied for most of the day. The house parrot confused us all with its devastatingly accurate imitation of Pat's voice: the creature had its very own television set to watch; when the programme did not suit its tastes it screamed to have the channel changed! And George, the estate factotum, provided endless cheerful assistance with lifting and shifting, and mild workshop stuff (we quickly forgot the rules about mangling parts: if it did not fit it was MADE to fit!). He would appear over the horizon just as you were looking for the fourth dropped nut on the bit that refused to go together with, “hi chaps! Are you winning?”

Every day, Pat would prepare home-made soup from the estate gardens and greenhouses. With fresh bread and grapes, figs and salads from the garden served on the terrace, we were in fine form.  Everything of course stopped for tea, when more of  Pat's home cooking would appear: we demolished whole cakes at a sitting so good were they. At the end of the day, Richard would appear, bearing armfuls of cucumbers, or tomatoes or figs, for us to take home.

I took a fair bit of “homework” away with me as I was falling behind schedule. Over two nights I built the traction chains and the main pinwheel drives for the Paternoster. However, the more of the thing I built, the less happy I was about the scale factor in the drives. Eventually I confided in Roger (who had returned from France by now, with us having a week or so to go for the St Albans deadline). Roger was positive: “let's get one chain track running and see”...  A trial with a half length of chain and a single pin drive ran, but I was still very concerned about how reliably the pin wheels would actually be on a fully loaded chain in practice: only one “missed” pin and the two chains would drop out of synchronisation, the cages would twist out of alignment and jam, and all hell would break loose. Once I had a full length of chain in its guides it was obvious that the frictional forces were far too high and the pin drives far too uncertain for anything approaching reliable running.

Meanwhile, Colin was playing with one of the new Parvalux 55 Watt mains motors we were using, to see how much gear reduction would be needed for his bit. He had lashed up a temporary take-off shaft directly driven from the motor armature when, suddenly, there was a clatter of flying parts and an expletive. Colin had neglected to lubricate the bearing of his shaft where it passed through a bushwheel boss. Within ten seconds, the powerful motor had friction-welded the shaft and the boss together completely irretrievably! Those were not the only parts to find their way into the scrap bin! But Colin persevered, and took the doubtful honour of finishing his part first.

Howard had built about half of the Carousel, but was now worried that the elegant pin drive for the main turntable, based on the drive for his famous Saxtead Mill, would not work. He had decided to hide the arrangement within the turntable bearing, so the drive pins pointed inwards, rather then vertically: there was a risk that the driving peg gear would foul the tips of the driven pins.

The other tower's framework was now in place so we discussed how to make a conventional lift mechanism inside it. Both Howard and I had made quite large prototype module constructions as homework, so Roger and Colin, with the occasional welcome help of Peter Harwood, could get on and build more or less unsupervised. The whole display was made up as much as possible from standard modular units, so one could just hand a prototype to a “gopher” (as we rudely christened those three) and tell them to make “n” off. We also made much use of powered screwdrivers, especially on large frameworks where everything is first assembled loosely and then progressively tightened to square. A tower module, for instance, used over 300 nuts and bolts on its main structure, so the electric screwdrivers were a real asset.

Roger designed  the “dunhill” trademark lettering in bright zinc and installed it at the base of each tower, which impressed Richard mightily!

Those towers took on a life of their own. Despite being built like the proverbial battleships, you could square and brace them all up nicely and return the next day to find them rhombohedral again! We discovered at this scale (eight feet high and two feet by eighteen inches in section) you simply could not have too much bracing.

The next hurdle was the Health and Safety clearance and the Dunhill electrician. Fortunately, I have had some experience of this from my time in manufacturing industry and Colin (now absent having his shoulder fixed) is something of an expert, and I also know a fair bit about electrics, but none of us had ever done anything quite like this.

We had decided that the Paternoster, the Carousel and Colin's Bridge would be driven continuously by the Parvalux motors we had been supplied with. Fortunately the armatures of these take a Meccano Socket Coupling very neatly. We took the precaution of using the modern Short version of this part to avoid shearing the Coupling at its narrow “neck”, and then coupled the motor up to Meccano shafts by belt (the Carousel), by constant velocity joint – a Handrail Coupling carrying a bolt located into the detent of a Socket Coupling (the Bridge), and by Universal Gears (the Paternoster – by far the heaviest load). The main safety issues were physically preventing people from touching any mechanisms – a child investigating the drive on the Carousel might easily loose a finger to the pinwheel – and getting power to the spotlights in each lift cage: a 20 Watt QI 12V bulb in each makes a potential current on the Paternoster of 15 Amps, which had to be picked up on sliding contacts from the chain guides. Fortunately this lighting current was AC and so we slipped under the regulations. Had it been DC, we would have been stymied, since even at that low voltage there are real risks from a high current DC source: you can get nasty burns and possible heart failure from such a power source. And one issue we did not expect; having avoided the low voltage DC problem we discovered that the height of the tower made the pick-up rails in the Paternoster ideal radio aerials, which could blank out all radio reception for a good radius around the store. We would not be popular!

When Bill, the store electrician, turned up he started by taking a look at the controller and low voltage drives on the conventional lift. I had already discussed these with him and he was able to provide us with a pair of lovely electronically-stabilised 12V power packs wired back to back to give us a split potential 24V power source. This proved perfect for our needs as the motors all ran absolutely predictably and stably, so the relatively crude timing arrangements in the controller stayed on track for days at a time. Once set up, the controller gave no trouble at all. We installed the massive toroidal lighting transformers in the bottom of the towers as useful ballast and wired all the mains back to a residual current protective device and a timer which would automatically start and stop the display for shop hours.

Slowly, things progressed. We found the St Alban's exhibition had crept up upon us all too soon, and the removers (Luke Wong) turned up to shift the assemblies up to the Show.  We decided to leave the half-built Carousel behind as we had not worked out how we could possible move it. In his quest for neatness, Howard had made its dismantling into easily carried sections almost impossible. Colin's beam was ready, but looked very spindly, the Paternoster remained an empty tower while I struggled to build a pair of traction wheels to drive the chain more positively than the pin wheels had done, but the conventional lift was a triumph. Howard built the controller and had built a fair prototype of the cage with its safety lock and guides, but Roger made it all work. It was a pretty close model of the real thing with a proper counterweight making the job of the winch motor easily done by a 12V DC motor. The only significant departure from prototype was the single leaf of doors at each landing: we were out to minimise complex machinery! Much to everyone's surprise, it worked the moment the controller was wired up. Not only did it work immediately, once the door mechanisms had settled down and the few rough spots on the guides had been sorted, it continued to run perfectly. Most encouraging.

Luke Wong had a big roll of bubble wrap, so we made a cushion of it in the back of his panel van, and gently lowered the twin towers onto it. The beam rode on top and a box of spares and most of a No 10 Set was loaded in the back. We met again at 07:00 the next morning at the Show. As Dave Taylor was supervising the setting of tables, we heaved and manoeuvred our stuff into place.

The conventional lift only needed its transport immobilisation removing and plugging in: it worked perfectly all day – though we did have to straighten a couple of bent girders and replace a dozen or so bolts that fell out during the move. However, the Beam shed a handful of nuts and bolts every time it was moved. Colin had been very economical with fastenings: over the weekend I added another 200 nuts and bolts to his frame, and Roger later beefed up the critical parts of the framework with extra girders, as single girders, in places where removal men are tempted to lift, got bent double in no time, we discovered. But it also worked without complaint all weekend.

The Paternoster stood immobile and silent as I continued to struggle with the design and then building the traction wheels: by the end of the Show I had built the two drive units and checked that the chains would run over them accurately. Much encouraged, we redoubled our efforts, back at Denham on Monday, to finish on time.

But by mid-week both Howard and I had seemingly intractable problems: one each.

The inward-facing drive pins on the Carousel's outer rotor did indeed prove to foul the driving pinwheel unpredictably. Turning the pins outward and mounting them visibly was clearly a good idea both mechanically and for maintenance, but the drive to them had to be re-thought, and a six-hole bushwheel was not going to work. Reluctantly, Howard ordered a pair of five-hole bushwheels from Lord Taylor of Dunmow, and set about re-re-re-building his main drive.

Meanwhile Roger and I had got slippery plastic guides mounted in angle girders to support the Paternoster's chains positively – thus, we hoped, avoiding any more complex guides for the cages – reassuringly exactly as per prototype. However, on mounting the chains to the traction wheels, massively supported in as stiff a framework as we could imagine, the axles just bent under the strain. The fundamental problem is that a Paternoster's drive must leave the space between the two hauling chains free for the cages to swing within. That, in turn, means that the drive to the traction wheels must be cantilevered out: the external bearing for each traction wheel is severely circumscribed by the very restricted clearance required between it and the passing cages. We doubled and redoubled strips and seriously considered moving to one of the large axle systems, but the framework simply was never going to stay in alignment at this scale and under these forces. Late that Thursday afternoon we resolved to scrap the Paternoster and replace it with what I characterised as a “see-saw” but which finally became known as the “Maternoster” - being the female derivative of the original and a “mother” of a mechanism to set up and look after!

As it turned out the “Maternoster” proved a better entertainer than  the Paternoster ever would have done. True of all things femail, I suppose  The design was based on the “Warehouse Lift” Supermodel, which offers an eccentric winch to haul the cages, thus automatically and mysteriously (from the beholder's point of view) slowing the cage down at each landing and then speeding them away each time. I built a pair, two separate mechanisms, so that the movements of the two cages would be independent and unpredictable.

Being a simple-minded Meccanoman, I had assumed that a standard reversing mechanism could be entrusted to change the direction of the winches at each end of the cages' travel. Wrong! Here we were dealing with very heavy loads, so high-torque drives with almost no overrun, to power a reversing mechanism past the neutral point. However I set it up, springs, toggles, extra trips, the reverser would neatly stop at the neutral point and refuse to engage the other direction. Finally I came up with the idea of tripping a very light reversing mechanism (two pinions and a contrate – you know the idea) with levers actuated by the winches or the cages (we settled on the latter for ease of setting up: given more time the former would have been more reliable), which then (being in constant mesh with the power source) drove the main reversing mechanism over. For this we used an all-pinion design to cope with the torque, which it did – perfectly. The tricky bit was to find a reliable slipping clutch for the reverser change-over drive – would you believe a simple rod connector with absolutely NO OIL?

By now we had run out of time. Roger – at the eleventh hour – negotiated a week's grace, and he, Howard and I worked almost around the clock to make it. We did – only just. I rode “shotgun” on Luke Wong's removal van that Sunday evening, very gently down to Jermyn Street. We had to park around the corner in Duke Street and carry the stuff from there. As promised, we stopped the traffic to do so. Dunhills had got the first part of our contract: we stopped the traffic!

Over the next week we installed the show and got it working safely. Although we had been reassured that screens and barriers would be in place, and a solid plinth ready for mounting, actually it transpired that Management was waiting to see what turned up before committing themselves to further expense! One could hardly blame them: half way through Tuesday, a Manager complained that “we were taking far too long setting up a window”. Ah! But they had never had a window like this. As John Abbate had promised, the sight of us scurrying around in the window attracted an almost permanent crowd of fascinated onlookers, some of whom came into the store. We had anticipated this and Dunhills had produced a beautifully printed coloured card (made up by our own Chris Bourne) to give away. The WLMS acknowledgement and website address was proudly displayed on the window itself. Slowly everyone got used to us, and even tolerated the permanent presence of a spare No 10 set parked behind the counter. By late November I was told we were “part of the Dunhill family now. We did not realise how close we would become to the West London Meccano Society” they said.

The finished models from the Duke Street window (Roger's Midland Compound and Terry's Napier) went to the Paris store for Christmas, but the Jermyn Street display carried on. “Happily ever after?” Not quite... Like stripping a pair of helical gears when the Maternoster jammed (as it did regularly); like reducing a pinion to a collar, like polishing a collar and a washer to a mirror flat finish, like powdering driving belts, like cutting clean through a pair of bushwheel bosses with an unlubricated axle rod. Every couple of days, someone would call Roger or me with “it's making a funny noise”, or “a noise like a machine gun, help”!

In mid-November, Dunhill threw a reception at the shop for everyone associated with the project and all the WLMS members too. Vintage champagne and elegant canapés kept us going through the speeches and exchanges of goodwill. In particular we learned that the store was not only doing better than most in the Street, but also it was doing much better than last year: perhaps on account of the endless stream of people stopping to watch the display and stepping, probably for their first time, into the shop itself. Finally Richard Dunhill himself delivered a brief word of thanks, telling us how much he had enjoyed being so close to the process. In return, Roger presented Pat with a bouquet, inadequate thanks for her wonderful hospitality.  [Left: Colin Davies, Howard Somerville, Paul Joachim, John Abbate (Dunhill), Roger Poulet, Richard Dunhill].

Would we do it again? Well, given a bigger challenge to go for, the answer has to be “at a price...” I would not do this one again, despite now knowing all the things we had not thought about when we started.

What did we get out of it? For myself, Dunhill's generous rewards and the glory of our names, and the club's name, in lights in the most prestigious part of London, are all secondary. What I got out of it was the sheer adrenaline rush from almost continuous Meccanoing against deadlines and in the company of splendid team mates. We worked ourselves to dropping point on several occasions, and even in those moments of exhausted despair we came through – after desperate brainstorming - with answers to the problems – engineering, logistical, safety, client-facing. The best bit for me was the amount of Meccano know-how I picked up just watching my colleagues building seriously difficult stuff. I know they in turn picked up some of my wheezes too. Mike Rhoades will benefit from selling another three sets of bolt-retaining screwdrivers, long bladed (“for making models in the next room”) and chunky screwdrivers. My magnetic pick-up stick disintegrated with all the use it got, and we used a whole packet of BluTak on our box spanners. Howard's forceps and Colin's trick spanners have also been copied now, and the secret for extended quiet running can also now be told: Castrol Hypoid EP140 oil for steel, and Castrol GTX Magnatec for brass: lubricants that insinuate themselves into everything and refuse to leave any part bare!

And the display? Well, in the New Year it was returned to Denham, where Richard Dunhill and faithful George spent the cold days of January and February slowly dismantling it, discovering its innermost secrets and returning the parts to the stock trays again. No different, really, from any other Meccano model.